Getting Set Up

Exploring the Data

Simple Plots in Base R

Plots with ggplot2

Getting Set Up

Before we begin, start a new file with File New File R Script. As you work through this sheet in the console in R, also add (copy/paste) your commands that work into this new file. At the end, save it, and run to execute all of your commands at once.

Exploring the Data

1.

We will look at GDP per Capita and Life Expectancy using some data from the gapminder project. There is a handy package called `gapminder` that uses a small snippet of this data for exploratory analysis. Install and load the package `gapminder`. Type `?gapminder` and hit enter to see a description of the data.

# first time only
# install.packages("gapminder")

# load gapminder
library(gapminder)

# get help
?gapminder

2

Let’s get a quick look at `gapminder` to see what we’re dealing with.

a.

Get the `str`ucture of the `gapminder` data.

str(gapminder)

## Classes 'tbl_df', 'tbl' and 'data.frame':    1704 obs. of  6 variables:
##  $ country  : Factor w/ 142 levels "Afghanistan",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ continent: Factor w/ 5 levels "Africa","Americas",..: 3 3 3 3 3 3 3 3 3 3 ...
##  $ year     : int  1952 1957 1962 1967 1972 1977 1982 1987 1992 1997 ...
##  $ lifeExp  : num  28.8 30.3 32 34 36.1 ...
##  $ pop      : int  8425333 9240934 10267083 11537966 13079460 14880372 12881816 13867957 16317921 22227415 ...
##  $ gdpPercap: num  779 821 853 836 740 ...

b.

What variables are there?

# - country: a factor
# - continent: a factor
# - year: an integer
# - lifeExp: a number
# - gdpPercap: a number

c.

Look at the `head` of the dataset to get an idea of what the data looks like.

head(gapminder)

ABCDEFGHIJ0123456789

country <fctr>	continent <fctr>	year <int>	lifeExp <dbl>	pop <int>	gdpPercap <dbl>
Afghanistan	Asia	1952	28.801	8425333	779.4453
Afghanistan	Asia	1957	30.332	9240934	820.8530
Afghanistan	Asia	1962	31.997	10267083	853.1007
Afghanistan	Asia	1967	34.020	11537966	836.1971
Afghanistan	Asia	1972	36.088	13079460	739.9811
Afghanistan	Asia	1977	38.438	14880372	786.1134

d.

Get `summary` statistics of all variables.

summary(gapminder)

##         country        continent        year         lifeExp     
##  Afghanistan:  12   Africa  :624   Min.   :1952   Min.   :23.60  
##  Albania    :  12   Americas:300   1st Qu.:1966   1st Qu.:48.20  
##  Algeria    :  12   Asia    :396   Median :1980   Median :60.71  
##  Angola     :  12   Europe  :360   Mean   :1980   Mean   :59.47  
##  Argentina  :  12   Oceania : 24   3rd Qu.:1993   3rd Qu.:70.85  
##  Australia  :  12                  Max.   :2007   Max.   :82.60  
##  (Other)    :1632                                                
##       pop              gdpPercap       
##  Min.   :6.001e+04   Min.   :   241.2  
##  1st Qu.:2.794e+06   1st Qu.:  1202.1  
##  Median :7.024e+06   Median :  3531.8  
##  Mean   :2.960e+07   Mean   :  7215.3  
##  3rd Qu.:1.959e+07   3rd Qu.:  9325.5  
##  Max.   :1.319e+09   Max.   :113523.1  
##

Simple Plots in Base R

3.

Let’s make sure you can do some basic plots before we get into the `gg`. Use base `R`’s `hist()` function to plot a histogram of `gdpPercap`.

hist(gapminder$gdpPercap)

4.

Use base `R`’s `boxplot()` function to plot a boxplot of `gdpPercap`.

boxplot(gapminder$gdpPercap)

5.

Now make it a boxplot by `continent`.¹

boxplot(gapminder$gdpPercap~gapminder$continent)

# alternate method
# boxplot(gdpPercap~continent, data = gapminder)

6.

Now make a scatterplot of `gdpPercap` on the -axis and `LifeExp` on the -axis.

plot(gapminder$lifeExp~gapminder$gdpPercap)

# alternate method
# boxplot(lifeExp~gdpPercap, data = gapminder)

Plots with `ggplot2`

7.

Load the package `ggplot2` (you should have installed it previously. If not, install first with `install.packages("ggplot2")`).

# install if you don't have
# install.packages("ggplot2")

# load ggplot2 
library(ggplot2)

8.

Let’s first make a `bar` graph to see how many countries are in each continent. The only `aes`thetic you need is to map `continent` to `x`. Bar graphs are great for representing categories, but not quantitative data.

ggplot(data = gapminder,
       aes(x = continent))+
  geom_bar()

9.

For quantitative data, we want a `histogram` to visualize the distribution of a variable. Make a `histogram` of `gdpPercap`. Your only `aes`thetic here is to map `gdpPercap` to `x`.

ggplot(data = gapminder,
       aes(x = gdpPercap))+
  geom_histogram()

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

10.

Now let’s try adding some color, specifically, add an `aes`thetic that maps `continent` to `fill`.²

ggplot(data = gapminder,
       aes(x = gdpPercap,
           fill = continent))+
  geom_histogram()

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

11.

Instead of a `histogram`, change the `geom` to make it a `density` graph. To avoid overplotting, add `alpha=0.4` to the `geom` argument (alpha changes the transparency of a `fill`).

ggplot(data = gapminder,
       aes(x = gdpPercap,
           fill = continent))+
  geom_density(alpha=0.4)

12.

Redo your plot from 11 for `lifeExp` instead of `gdpPercap`.

ggplot(data = gapminder,
       aes(x = lifeExp,
           fill = continent))+
  geom_density(alpha=0.4)

13.

Now let’s try a scatterplot for `lifeExp` (as `y`) on `gdpPercap` (as `x`). You’ll need both for `aes`thetics. The `geom` here is `geom_point()`.

ggplot(data = gapminder,
       aes(x = gdpPercap,
           y = lifeExp))+
  geom_point()

14.

Add some color by mapping `continent` to `color` in your `aes`thetics.

ggplot(data = gapminder,
       aes(x = gdpPercap,
           y = lifeExp,
           color = continent))+
  geom_point()

15.

Now let’s try adding a regression line with `geom_smooth()`. Add this layer on top of your `geom_point()` layer.

ggplot(data = gapminder,
       aes(x = gdpPercap,
           y = lifeExp,
           color = continent))+
  geom_point()+
  geom_smooth()

## `geom_smooth()` using method = 'loess' and formula 'y ~ x'

16.

Did you notice that you got multiple regression lines (colored by continent)? That’s because we set a `global` `aes`thetic of mapping `continent` to `color`. If we want just one regression line, we need to instead move the `color = continent` inside the `aes` of `geom_point`. This will only map `continent` to `color` for points, not for anything else.

ggplot(data = gapminder,
       aes(x = gdpPercap,
           y = lifeExp))+
  geom_point(aes(color = continent))+
  geom_smooth()

## `geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'

17.

Now add an `aes`thetic to your `point`s to map `pop` to `size`.

ggplot(data = gapminder,
       aes(x = gdpPercap,
           y = lifeExp))+
  geom_point(aes(color = continent, 
                 size = pop))+
  geom_smooth()

## `geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'

18.

Change the color of the regression line to `"black"`. Try first by putting this inside an `aes()` in your `geom_smooth`, and try a second time by just putting it inside `geom_smooth` without an `aes()`. What’s the difference, and why?

ggplot(data = gapminder,
       aes(x = gdpPercap,
           y = lifeExp))+
  geom_point(aes(color = continent, 
                 size = pop))+
  geom_smooth(aes(color = "black"))

## `geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'

# putting it inside aesthetics tries to map color to something
# in the da ta called "black", since R can't find "black", 
# it will produce some random color

ggplot(data = gapminder,
       aes(x = gdpPercap,
           y = lifeExp))+
  geom_point(aes(color = continent, 
                 size = pop))+
  geom_smooth(color = "black")

## `geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'

# putting it outside aesthetics (correctly) sets color to black

19.

Another way to separate out continents is with `facet`ing. Add `+facet_wrap(~continent)` to create subplots by `continent`.

ggplot(data = gapminder,
       aes(x = gdpPercap,
           y = lifeExp))+
  geom_point(aes(color = continent, 
                 size = pop))+
  geom_smooth(color = "black")+
  facet_wrap(~continent)

## `geom_smooth()` using method = 'loess' and formula 'y ~ x'

20.

Remove the `facet` layer. The `scale` is quite annoying for the `x`-axis, a lot of points are clustered on the lower level. Let’s try changing the scale by adding a layer: `+scale_x_log10()`.

ggplot(data = gapminder,
       aes(x = gdpPercap,
           y = lifeExp))+
  geom_point(aes(color = continent, 
                 size = pop))+
  geom_smooth(color="black")+
  scale_x_log10()

## `geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'

21.

Now let’s fix the labels by adding `+labs()`. Inside `labs`, make proper axes titles for `x`, `y`, and a `title` to the plot. If you want to change the name of the legends (continent color), add one for `color` and `size`.

ggplot(data = gapminder,
       aes(x = gdpPercap,
           y = lifeExp))+
  geom_point(aes(color = continent, 
                 size = pop))+
  geom_smooth(color="black")+
  scale_x_log10()+
  labs(x = "GDP per Capita",
       y = "Life Expectancy",
       color = "Continent",
       size = "Population")

## `geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'

22.

Now let’s try subsetting by looking only at North America. Take the `gapminder` dataframe and subset it to only look at `continent=="Americas"`). Assign this to a new dataframe object (call it something like `america`.) Now, use this as your `data`, and redo the graph from question 17. (You might want to take a look at your new dataframe to make sure it worked first!)

america<-gapminder[gapminder$continent=="Americas",]

# verify this worked
america

ABCDEFGHIJ0123456789

country <fctr>	continent <fctr>	year <int>	lifeExp <dbl>	pop <int>	gdpPercap <dbl>
Argentina	Americas	1952	62.485	17876956	5911.315
Argentina	Americas	1957	64.399	19610538	6856.856
Argentina	Americas	1962	65.142	21283783	7133.166
Argentina	Americas	1967	65.634	22934225	8052.953
Argentina	Americas	1972	67.065	24779799	9443.039
Argentina	Americas	1977	68.481	26983828	10079.027
Argentina	Americas	1982	69.942	29341374	8997.897
Argentina	Americas	1987	70.774	31620918	9139.671
Argentina	Americas	1992	71.868	33958947	9308.419
Argentina	Americas	1997	73.275	36203463	10967.282

ggplot(data = america,
       aes(x = gdpPercap,
           y = lifeExp))+
  geom_point(aes(color = continent, 
                 size = pop))+
  geom_smooth()

## `geom_smooth()` using method = 'loess' and formula 'y ~ x'

23.

Try this again for the whole world, but just for observations in the year 2002.

gap_2002<-gapminder[gapminder$year==2002,]

# verify this worked
gap_2002

ABCDEFGHIJ0123456789

country <fctr>	continent <fctr>	year <int>	lifeExp <dbl>	pop <int>	gdpPercap <dbl>
Afghanistan	Asia	2002	42.129	25268405	726.7341
Albania	Europe	2002	75.651	3508512	4604.2117
Algeria	Africa	2002	70.994	31287142	5288.0404
Angola	Africa	2002	41.003	10866106	2773.2873
Argentina	Americas	2002	74.340	38331121	8797.6407
Australia	Oceania	2002	80.370	19546792	30687.7547
Austria	Europe	2002	78.980	8148312	32417.6077
Bahrain	Asia	2002	74.795	656397	23403.5593
Bangladesh	Asia	2002	62.013	135656790	1136.3904
Belgium	Europe	2002	78.320	10311970	30485.8838

ggplot(data = gap_2002,
       aes(x = gdpPercap,
           y = lifeExp))+
  geom_point(aes(color = continent, 
                 size = pop))+
  geom_smooth()

## `geom_smooth()` using method = 'loess' and formula 'y ~ x'

Hint: use formula notation with~.↩︎
In general, color refers to the outside borders of a geom (except points), fill is the interior of an object.↩︎

1.4 Data Visualization with ggplot2 - R Practice (Answers)

Ryan Safner

ECON 480 - Fall 2019

Getting Set Up

Exploring the Data

1.

2

Let’s get a quick look at gapminder to see what we’re dealing with.

a.

Get the structure of the gapminder data.

b.

What variables are there?

c.

Look at the head of the dataset to get an idea of what the data looks like.

d.

Get summary statistics of all variables.

Simple Plots in Base R

3.

Let’s make sure you can do some basic plots before we get into the gg. Use base R’s hist() function to plot a histogram of gdpPercap.

4.

Use base R’s boxplot() function to plot a boxplot of gdpPercap.

5.

Now make it a boxplot by continent.1

6.

Now make a scatterplot of gdpPercap on the x-axis and LifeExp on the y-axis.

Plots with ggplot2

7.

Load the package ggplot2 (you should have installed it previously. If not, install first with install.packages("ggplot2")).

8.

Let’s first make a bar graph to see how many countries are in each continent. The only aesthetic you need is to map continent to x. Bar graphs are great for representing categories, but not quantitative data.

9.

For quantitative data, we want a histogram to visualize the distribution of a variable. Make a histogram of gdpPercap. Your only aesthetic here is to map gdpPercap to x.

10.

Now let’s try adding some color, specifically, add an aesthetic that maps continent to fill.2

11.

Instead of a histogram, change the geom to make it a density graph. To avoid overplotting, add alpha=0.4 to the geom argument (alpha changes the transparency of a fill).

12.

Redo your plot from 11 for lifeExp instead of gdpPercap.

13.

Now let’s try a scatterplot for lifeExp (as y) on gdpPercap (as x). You’ll need both for aesthetics. The geom here is geom_point().

14.

Add some color by mapping continent to color in your aesthetics.

15.

Now let’s try adding a regression line with geom_smooth(). Add this layer on top of your geom_point() layer.

16.

17.

Now add an aesthetic to your points to map pop to size.

18.

Change the color of the regression line to "black". Try first by putting this inside an aes() in your geom_smooth, and try a second time by just putting it inside geom_smooth without an aes(). What’s the difference, and why?

19.

Another way to separate out continents is with faceting. Add +facet_wrap(~continent) to create subplots by continent.

20.

Remove the facet layer. The scale is quite annoying for the x-axis, a lot of points are clustered on the lower level. Let’s try changing the scale by adding a layer: +scale_x_log10().

21.

Now let’s fix the labels by adding +labs(). Inside labs, make proper axes titles for x, y, and a title to the plot. If you want to change the name of the legends (continent color), add one for color and size.

22.

23.

Try this again for the whole world, but just for observations in the year 2002.

Let’s get a quick look at `gapminder` to see what we’re dealing with.

Get the `str`ucture of the `gapminder` data.

Look at the `head` of the dataset to get an idea of what the data looks like.

Get `summary` statistics of all variables.

Let’s make sure you can do some basic plots before we get into the `gg`. Use base `R`’s `hist()` function to plot a histogram of `gdpPercap`.

Use base `R`’s `boxplot()` function to plot a boxplot of `gdpPercap`.

Now make it a boxplot by `continent`.¹

Now make a scatterplot of `gdpPercap` on the -axis and `LifeExp` on the -axis.

Plots with `ggplot2`

Load the package `ggplot2` (you should have installed it previously. If not, install first with `install.packages("ggplot2")`).

Let’s first make a `bar` graph to see how many countries are in each continent. The only `aes`thetic you need is to map `continent` to `x`. Bar graphs are great for representing categories, but not quantitative data.

For quantitative data, we want a `histogram` to visualize the distribution of a variable. Make a `histogram` of `gdpPercap`. Your only `aes`thetic here is to map `gdpPercap` to `x`.

Now let’s try adding some color, specifically, add an `aes`thetic that maps `continent` to `fill`.²

Instead of a `histogram`, change the `geom` to make it a `density` graph. To avoid overplotting, add `alpha=0.4` to the `geom` argument (alpha changes the transparency of a `fill`).

Redo your plot from 11 for `lifeExp` instead of `gdpPercap`.

Now let’s try a scatterplot for `lifeExp` (as `y`) on `gdpPercap` (as `x`). You’ll need both for `aes`thetics. The `geom` here is `geom_point()`.

Add some color by mapping `continent` to `color` in your `aes`thetics.

Now let’s try adding a regression line with `geom_smooth()`. Add this layer on top of your `geom_point()` layer.

Now add an `aes`thetic to your `point`s to map `pop` to `size`.

Change the color of the regression line to `"black"`. Try first by putting this inside an `aes()` in your `geom_smooth`, and try a second time by just putting it inside `geom_smooth` without an `aes()`. What’s the difference, and why?

Another way to separate out continents is with `facet`ing. Add `+facet_wrap(~continent)` to create subplots by `continent`.

Remove the `facet` layer. The `scale` is quite annoying for the `x`-axis, a lot of points are clustered on the lower level. Let’s try changing the scale by adding a layer: `+scale_x_log10()`.

Now let’s fix the labels by adding `+labs()`. Inside `labs`, make proper axes titles for `x`, `y`, and a `title` to the plot. If you want to change the name of the legends (continent color), add one for `color` and `size`.